Soil fungi play crucial roles in decomposition and plant nutrient uptake, which are important ecosystem processes in restoring biodiversity. Understanding how soil fungal communities reassemble on degraded lands could help us enhance systems to reach restoration targets. Here we studied fungal community structures in a restoration area in cool-temperate forests of northern Japan. Our study used a two-way factorial design with eight treatment combinations: vegetation types (grassland, larch plantation, mixed plantation, and natural forest) × deer fence (exclosure and control). Larches and mixtures of conifers and broadleaves were planted a few decades ago to restore former arable lands. Grasslands were sites where such reforestation practices have failed (i.e. grassland = negative control, natural forest = positive control). Deer fences were built more than a decade ago to enhance the recovery of forest floor vegetation and tree regeneration. We measured fungal community structures by means of DNA metabarcoding in nine plots within each treatment combination (total n=72). Differences in diversity (alpha, beta, and gamma) and species composition among treatments were analyzed.
Mean alpha diversity (fungal OTU richness in each plot) was lower in natural forests than in the other three vegetation types. There were no significant differences among the rest of the vegetation types, nor between the inside and outside of the deer fences. Gamma diversity (total fungal OTU richness in each treatment) was also lower in natural forests than in other vegetation types. Deer fences again had no influence. Two-way PERMANOVA revealed that vegetation types, but not deer fences, had significant effects on fungal community composition. Guild-level analyses showed that coprophilous fungi occurred frequently in grasslands and outside deer fences, which is attributable to the fact that deer spend a long time in grasslands (especially during winter) and are present only outside the fences. Beta diversity (among-treatment difference in the size of within-treatment community dissimilarity) was significantly higher in natural forests than in other vegetation types. Overall, despite relatively long and intensive restoration practices being conducted in our study area, fungal diversity in afforested sites (larch and mixed plantations) were more similar to that in grasslands than natural forests, and deer fences had limited influence. Further progress in restoration schemes as well as conservation of natural forests are needed.